An in vitro assessment of the response of THP-1 macrophages to varying stiffness of a glycol-chitosan hydrogel for vocal fold tissue engineering applications

Abstract

The physical properties of a biomaterial play an essential role in regulating immune and reparative activities within the host tissue. This study aimed to evaluate the immunological impact of material stiffness of a glycol-chitosan hydrogel designed for vocal fold tissue engineering. Hydrogel stiffness was varied via the concentration of glyoxal cross-linker applied. Hydrogel mechanical properties were characterized through atomic force microscopy and shear plate rheometry. Using a transwell setup, macrophages were co-cultured with human vocal fold fibroblasts that were embedded within the hydrogel. Macrophage viability and cytokine secretion were evaluated at 3, 24, and 72 hr of culture. Flow cytometry was applied to evaluate macrophage cell surface markers after 72 hr of cell culture. Results indicated that increasing hydrogel stiffness was associated with increased anti-inflammatory activity compared to relevant controls. In addition, increased anti-inflammatory activity was observed in hydrogel co-cultures. This study highlighted the importance of hydrogel stiffness from an immunological viewpoint when designing novel vocal fold hydrogels.

1 |. INTRODUCTION

Hydrogels are water-swollen polymer networks that have been engineered for biomaterial therapeutics of defective tissues and organs.^1–4 There is growing interest in tuning the biomechanical properties of hydrogels to facilitate favorable tissue regeneration outcomes.^5–7 For example, hydrogel stiffness can affect focal adhesion maturation and the levels of tyrosine phosphorylation of FAK and paxillin expression.^8,9 In addition, porosity can significantly affect the onset and severity of chronic inflammation.¹⁰

Various methods have been reported for tuning hydrogel properties for desired biological outcomes. For instance, the degradation rate of PEG-based hydrogels can be modified by introducing α-hydroxy acids or enzymatically cleavable peptides to the PEG macromer backbone.^11,12 Furthermore, hydrogel stiffness can be modified by varying cross-linking density present via, for example, altering the concentration of cross-linker applied.^13,14

The immunomodulatory properties of vocal fold hydrogels have been explored using vocal fold fibroblasts (VFFs), macrophages (Mφ), mesenchymal stem cells, and adipose-derived stem cells.^15–21 Specifically, VFFs represent the most abundant cell type found within the lamina propria where they serve to maintain and repair the vocal fold mucosa.^22–24 VFFs have been recognized for their mechanosensitive and immunomodulatory properties including their ability to suppress pro-inflammatory cytokine production.^25–28 Fibroblast-like cells have been suggested to have some behavioral capacities similar to innate immune cells and also play a role in inflammatory resolution.²⁹ Furthermore, VFFs influence the local tissue environment through the provision of glycosaminoglycans (GAGs), proteogylcans, elastin, and collagen that help dictate the migration, growth, and differentiation of surrounding cells.^22,25,30

Hydrogel physical properties have been demonstrated to influence VFF activity in vitro. For example, decreased cross-linking density of PEGDA hydrogels was shown to induce increased GAG synthesis and decreased elastin synthesis from VFFs.³¹ HA-gelatin composite microgels demonstrated that other physical properties, such as a controlled degradation rate and increased surface area, can enhance VFF behavior including their adhesion capacity and motility.³²

In contrast, Mφ are antigen-presenting immune cells that are one of the early cellular responders to an implanted hydrogel. Critically, Mφ dysregulation can induce excessive fibrosis of the local tissue.³³ Hydrogel physical properties have been associated with alterations in Mφ phenotype alongside changes in chemokine, cytokine, and matrix metalloproteinase production.^34–36 For example, the pore size of poly (2-hydroxyethyl methacrylate-methacrylic acid) hydrogels was associated with a shift in Mφ phenotype and associated inflammatory activity.³⁷ Pore sizes >20 μm induced a shift to an M2 phenotype whereas nonporous hydrogels were related to an increased M1 population.

Co-cultures of VFFs and Mφ have been applied to quantify immunomodulatory properties of vocal fold hydrogels.¹⁸ Evidence of crosstalk between VFFs and Mφ through paracrine signaling has been demonstrated.³⁸ For instance, Mφ secrete IL-10 following a shift from an “inflammatory” to a “tissue repair” phenotype.^34,35,39 IL-10 presence has been associated with promoting healthy, nonfibrotic VFFs in PEGDA hydrogels.¹⁹ In addition, fibroblasts stimulated by IL-10 have been shown to synthesize ECM components essential to reestablishing normal tissue structure and function.⁴⁰ These findings further support the idea that the immunomodulatory properties of vocal fold hydrogels can be exploited to promote inflammatory resolution and ECM remodeling for effective tissue repair.

Stiffness is a notable physical design parameter of vocal fold hydrogels because of the unique elastic and stiffness properties human vocal fold tissue requires for phonation. Depending upon age and gender, human vocal folds typically vibrate at frequencies in the range of 100–300 Hz during phonation.^41–44 The Young’s moduli, or stiffness, of human vocal folds ranges from 2.45 to 29.4 kPa.⁴⁵ If vocal fold hydrogel stiffness is not tuned to match physiological standards, phonatory function will not be properly restored. More importantly, in addition to the importance of vocal fold hydrogel stiffness for replicating the native tissue environment, stiffness offers an opportunity to modulate the innate immune response.

Cells sense and respond to the surrounding local matrix through mechanotransduction.^46,47 Interactions between a biomaterial and the local tissue environment have been shown to regulate cellular behavior including contractibility, differentiation, motility, and spreading.^48–51 In particular, local cells can sense the material stiffness and trigger specific intracellular signaling events. Material stiffness has been demonstrated to influence focal adhesion formation and maturation.^8,9 Soft hydrogels (0.1–1 kPa) can limit focal adhesion maturation and increase their turnover rate. Stiffer hydrogels (10–100 kPa) can produce extended focal adhesions with increased levels of tyrosine phosphorylation of FAK and paxillin expression. Cytoskeleton assembly can also be modulated through material stiffness. For example, fibroblasts cultured on soft polyacrylamide gels (~0.2 kPa) displayed a spherical appearance with no stress fibers, whereas on stiffer gels (~3.6 kPa) fibroblasts were oblong with a visible cytoskeleton.⁵²

A mechanically tunable glycol-chitosan hydrogel has been proposed for vocal fold tissue engineering.¹⁴ The concentration of glyoxal, a dialdehyde cross-linker, was used to modify the stiffness of chitosan-based hydrogels.⁵³ Higher glyoxal concentrations produce stiffness values more closely aligned with human vocal folds that have been reported to have an average Young’s modulus of 12.66 kPa.⁴⁵ Unfortunately, increased glyoxal content within glycol-chitosan hydrogels has previously been associated with reduced VFF viability.¹⁴ As such, an appropriate glyoxal concentration must be selected that provides the hydrogel with mechanical integrity and relevant stiffness while minimizing cytotoxic effects upon cells.^54,55

However, VFF hydrogel monocultures provide only limited information on the nature of the innate inflammatory response associated with glycol-chitosan hydrogels. Further investigation is necessary to quantify inflammatory effects related to varying the glyoxal concentration of the hydrogels. Given the significance of Mφ in biomaterial-related inflammatory activity, Mφ inclusion in a subsequent in vitro model will enable the inflammatory response to be quantified more extensively.⁵⁶

The primary purpose of this study was to investigate the response of human Mφ to stiffness variations of a glycol-chitosan hydrogel in the presence of human VFFs. Mechanical properties of the hydrogel were characterized with atomic force microscopy (AFM) and oscillatory shear rheometry. Mφ inflammatory activity was characterized over a period of 72 hr via assessments of cytokine production, cell viability, and cell phenotyping. The hypothesis was that stiffer hydrogels would be associated with increased Mφ pro-inflammatory activity over a 72 hr period due to the cytotoxicity of the glyoxal cross-linker.

2 |. MATERIALS AND METHODS

2.1 |. Glycol-chitosan hydrogel preparation

A stock solution of glycol-chitosan hydrogel was prepared by mixing 4 g of glycol-chitosan powder (Sigma-Aldrich, Lot #SLBV0662) with 80 ml of 1 × phosphate buffered saline (PBS) (Wisent Bioproducts, QC, Canada) at room temperature to form a glycol-chitosan solution with a concentration of 4.76% (wt/wt %). The solution was rotated (MX-RDPro, Montreal Biotech Inc.) at 10 rpm for 24 hr. Following rotation, the solution was sterilized by autoclaving (SV-120, Steris Amsco Century) for 30 min at 121°C before being stored at 4°C until use.

2.2 |. Glyoxal cross-linker preparation

A stock solution of 10% glyoxal cross-linker solution was prepared in a fume hood at room temperature by mixing 5 ml of 40% glyoxal solution (Sigma-Aldrich, Lot #MKBV9169V) with 15 ml 1 × PBS. The 10% glyoxal solution was then sterilized by autoclaving and stored at 4°C. Prior to each experiment, 10% glyoxal solution was mixed with sterile PBS in a biosafety cabinet to produce three glyoxal solutions with concentration 0.005, 0.01, and 0.02% respectively.

2.3 |. Mechanical characterization–oscillatory shear rheometry

A TA Instrument Rheometer-AR2000 (New Castle, DE) was used to determine the storage and loss moduli of the cross-linked glycol-chitosan hydrogel. Parallel plates with a diameter of 20 mm and gap length 500 μm were used for testing and the plate temperature was set to 37°C. In a test tube, 400 μl of the glycol-chitosan hydrogel was mixed with 500 μl of defined glyoxal cross-linker (0.005, 0.01, or 0.02%). 300 μl of the resultant mix was applied to the bottom rheometer plate. 200 μl of baby oil (Johnson & Johnson) was applied around the plate edge to prevent sample dehydration. Soak time was set to 2 hr to allow hydrogel gelation prior to testing. Following soak time, a controlled stress frequency sweep test was performed between the angular frequency range 0.1–100 rad/s.

2.4 |. Mechanical characterization–AFM

AFM was used to measure substrate mechanical properties of hydrogels on cellular scale.⁴ Cross-linked hydrogel samples were prepared in 35 mm glass bottomed petri dishes (MatTek Corporation, USA) by mixing 400 μl of the glycol-chitosan hydrogel with 500 μl of defined glyoxal cross-linker (0.005, 0.01, or 0.02%). The hydrogel was distributed using a cell scraper to form a thin film across the dish surface. Each hydrogel sample was placed in an incubator (37°C, 5% CO₂) for 2 hr to facilitate gelation. Following gelation, PBS was added atop the hydrogel to prevent dehydration. AFM (JPK Nano-wizard 3, Bio-Science, Berlin, Germany) was then applied to each hydrogel to determine the Young’s modulus of each respective cross-linked hydrogel. A rectangular silicon cantilever (0.6 N/m) with an attached 25 μm polystyrene spherical particle (Novascan Technologies, Inc., Boone, IA) was applied for all testing. 20 unique points on each hydrogel surface were tested and results were averaged to obtain the mean Young’s modulus of each hydrogel.

2.5 |. THP-1 monocyte culture

The human monocyte cell line THP-1 (ATCC TIB202, Manassas, VA) was maintained in suspension in Roswell Park Memorial Institute medium (RPMI) 1640 (Wisent Bioproducts, QC, Canada) supplemented with 10% (vol/vol) fetal bovine serum (FBS) (Wisent Bioproducts, QC, Canada) and 1% (vol/vol) penicillin streptomycin (Wisent Bioproducts, QC, Canada). Cell cultures were stored in a humidified incubator (37°C, 5% CO₂) for amplification.

2.6 |. Induced differentiation of THP-1 monocytes

To obtain THP-1 Mφ, 5 ml of THP-1 monocytes at a concentration of approximately 1 × 10⁶ cells/ml were transferred to a tissue culture treated cell culture dish (100 × 20 mm, Eppendorf).⁵⁷ Phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich) was added to the cells at a concentration of 20 ng/ml. Cells were incubated for 24 hr to allow differentiation. A light microscope (Axiovert 40C, Zeiss) was used to confirm differentiation visualized by cell adherence to the petri dish surface and morphological changes. Phenotypical changes were verified with flow cytometry as described further below.⁵⁸

2.7 |. VFF culture

Immortalized human VFFs donated by the Thibeault laboratory (University of Wisconsin-Madison)⁵⁹ were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Wisent Bioproducts, QC, Canada) supplemented with 10% (vol/vol) FBS, 1% (vol/vol) penicillin streptomycin and 1% (vol/vol) nonessential amino acid (Sigma-Aldrich). VFF cultures were stored in a humidified incubator (37°C, 5% CO₂). At approximately 90% confluency, media was removed and cells were washed with PBS before TrypLE (Gibco) was added for 5 min to facilitate cell-surface detachment. A light microscope was used to confirm cell detachment. Supplemented DMEM was added to stop dissociation. Cells were counted using trypan blue (Invitrogen) and a hemocytometer (Fisher Scientific, Cat #0267110) before being centrifuged at 900 rpm for 5 min. Media was removed and cells were resuspended in supplemented phenol red-free DMEM at a working volume of approximately 1 × 10⁶ cells/ml. VFFs with a passage number of 6–10 were used throughout this study.

2.8 |. Transwell cell culture

Four testing groups were prepared for transwell culture (Table 1).

TABLE 1.

Testing groups applied for transwell culture

Culture	Group	Label	Setup
Hydrogel co-culture of THP-1 Mφ and VFFs	Experimental	(Mφ + VFF + hydrogel)	Mφ seeded on basolateral transwell membrane, VFFs embedded in hydrogel in upper well
Hydrogel monoculture of THP-1 Mφ	Control	(Mφ + hydrogel)	Mφ seeded on basolateral transwell membrane, acellular hydrogel in upper well
Nonhydrogel co-culture of THP-1 Mφ and VFFs	Control	(Mφ + VFF)	Mφ seeded on basolateral transwell membrane, VFFs seeded on apical transwell membrane
Nonhydrogel monoculture of THP-1 Mφ	Control	(Mφ)	Mφ seeded on basolateral transwell membrane

Abbreviations: Mφ, macrophages; VFFs, vocal fold fibroblasts.

To prepare PMA-induced Mφ for experiments, media and non-adherent cells were aspirated from the petri dish. Mφ were then washed with 5 ml 1 × PBS before 2 ml of TrypLE (Fisher Scientific, Ottawa, ON, Canada) was applied. The dish was placed in an incubator (37°C, 5% CO₂) for 5 min to facilitate dissociation. Cells were checked under a light microscope for nonadherence before 5 ml of supplemented phenol-red free RMPI (Wisent Bioproducts, QC, Canada) was added to stop dissociation. Cells were transferred to a 50 ml polypropylene centrifuge tube and the dish surface was washed with an additional 5 ml of RPMI to collect remaining cells. Cells were stained with trypan blue and counted using a cell hemocytometer before being centrifuged at 900 rpm for 5 min. Media was aspirated and the cells resuspended in supplemented phenol red-free RPMI at a working volume of approximately 1.5 × 10⁶ cells/ml.

Tissue culture treated 6-well transwell plates (polycarbonate membrane, 0.4 μm pore, Corning, Cat #3412) were used for all cell cultures (Figure 1). Transwell inserts enabled the physical separation of VFFs and Mφ while permitting cytokine transport between cell types, mimicking paracrine signaling.⁶⁰ 500 μl of Mφ were seeded on the basolateral underside of a tissue culture-treated transwell and placed in an incubator (37°C, 5% CO₂) top-side down for 3 hr. After incubation, transwells plates were gently turned over (top-side up) and 1 ml of supplemented phenol-red free RPMI media was added to lower wells. In the upper well of (Mφ + VFF + hydrogel) cultures, 400 μl of glycol-chitosan hydrogel was mixed with 500 μl of glyoxal cross-linker (0.005, 0.01, or 0.02%) and 200 μl of HVFF cells. For (Mφ + hydrogel) cultures, 200 μl of DMEM media was applied instead of VFFs. Finally, (Mφ + VFF) and (Mφ) cultures contained no hydrogel in the upper well and instead applied only 200 μl of VFFs ([Mφ + VFF]) or DMEM media ([Mφ]) in the upper well. Transwell plates were stored in an incubator for 2 hr to permit hydrogel gelation. After 2 hr, plates 750 μl of supplemented DMEM was added to all transwell upper wells to prevent hydrogel dehydration and provide VFFs with additional nutrients. Plates were then incubated for 3, 24, or 72 hr.

FIGURE 1.

Transwell setup for the ([Mφ + VFF + hydrogel]) culture. Mφ were seeded on the basolateral transwell membrane and VFFs were embedded in the hydrogel. For (Mφ + hydrogel), VFFs were excluded. For (Mφ + VFF), the hydrogel was excluded. For (Mφ), VFFs, and the hydrogel were excluded. DMEM, Dulbecco’s modified Eagle’s medium; Mφ, macrophages; VFFs, vocal fold fibroblasts

Supernatant samples from the lower well, representative of the 3, 24, and 72 hr Mφ response, were collected and stored at −80°C prior to analysis. Immediately following supernatant collection, Mφ on the basolateral transwell membrane was stained for viability analysis. Separate culture plates were used for each time-point to ensure supernatants collected were cumulative over the defined time period for cytokine protein analyses. For Mφ phenotyping via flow cytometry, only 72 hr culture samples were examined. All experiments were performed in at least triplicate.

2.9 |. Secreted pro- and anti-inflammatory cytokine levels

Enzyme-linked immunosorbent assays (ELISAs) were used to measure the concentration of pro- and anti-inflammatory cytokines in supernatant samples. Human tumor necrosis factor alpha (TNF-α) (Cat. #KHC3011, Invitrogen, Vienna, Austria) and human interleukin 10 (IL-10) (Cat. #HS100C, R&D systems, MN) were assessed to determine the respective pro- and anti-inflammatory response. Standard curves were applied to measure the quantity of analyte present in each sample. After thawing samples on ice, vials were briefly centrifuged (1,200 rpm) to remove debris prior to loading into the respective ELISA plates. Total protein concentration was quantified using the Bradford Protein Assay (Cat. #5000002, Bio-Rad Laboratories, Canada) and used to normalize ELISA data. All assays were performed according to the manufacturer’s requirements.

2.10 |. Cell viability

The Mφ seeded on the basolateral underside of the transwell membrane were washed with 1 × PBS before being stained using a LIVE/DEAD viability/cytotoxicity assay (Invitrogen; Eugene, OR, Lot #1932445) in accordance with the manufacturer’s instructions. The plate was incubated for 30 min in darkness at room temperature before subsequently being washed twice more with PBS. The transwell insert was then placed into 35 mm glass bottom culture dishes (MatTek Corporation) containing 250 μl of fresh PBS. An inverted confocal fluorescence microscope (LSM710, Zeiss, Germany) was used to image the stained samples. Images were acquired using a 10× objective with channels of 488 nm (LIVE, green channel) and 514 nm (DEAD, red channel).

Image acquisition and analysis were performed using Zen System (Zeiss, Germany) and Imaris version 7.5.6 (Bitplane, South Windsor, CT) software. Mφ viability (%) was determined by dividing the number of live Mφ by the total Mφ count (live + dead) and multiplying the result by 100. A colocalization channel was applied to account for cells that fluoresced both green and red; such cells were counted as dead for viability calculations.

2.11 |. Cell phenotyping

Four-color flow cytometry was conducted to assess how the glycol-chitosan hydrogel influenced Mφ phenotype. To harvest Mφ for flow cytometry, RPMI media was aspirated from each lower well and 1.2 ml of PBS was added for washing. PBS was then aspirated and 1 ml of TrypLE was added before the plate was placed in an incubator (37°C, 5% CO₂) for 5 min to facilitate dissociation. Following incubation the plate was tapped to promote further cell detachment before 1.2 ml of supplemented RPMI was added to each well to stop dissociation. Cells were then collected and transferred to a 50 ml tube containing 5 ml of supplemented RPMI. The basolateral membrane of the transwell was washed with a further 1 ml of RPMI to collect any remaining Mφ that were subsequently added to the 50 mL tube. Cells were centrifuged at 900 rpm before being resuspended in 10 ml of PBS and centrifuged again at 900 rpm for washing. Finally, cells were resuspended at a working volume of 1 × 10^6 cells/ml in sodium azide- and protein-free 1 × DPBS. 150 μl of cell solution was transferred to each labeled flow cytometry tube (BD Biosciences) and 5 μl of Human BD Fc Block (BD #564219) was added per tube followed by 10 min incubation at room temperature. A preprepared antibody cocktail was added to each cell aliquot suspension. The cocktail comprised 50 μl of Brilliant Stain Buffer (BSB) (BD #563794) and 5 μl of each of the four markers. The four-marker panel constructed for this experiment consisted of CD11b (#562721), CD33 (#564588), CD80 (#560925), and CD206 (#564063) (BD Life Sciences, San Jose, CA) (Table 1). CD11b and CD33 identified nonpolarized, basal Mφ⁵⁸ and confirmed monocyte to Mφ differentiation. CD80 identified pro-inflammatory Mφ⁶¹. CD206 identified anti-inflammatory Mφ.⁶¹ (Mφ + VFF + hydrogel) cultures were compared to (Mφ + hydrogel) cultures, and (Mφ) cultures.

Samples were prepared as unstained controls (50 μl BSB only), single stained controls (50 μl BSB + 5 μl of one antibody), fluorescence minus one (FMO) controls (50 μl BSB + 5 μl of three of the four antibodies), or multi-stained samples (50 μl BSB + 5 μl of all antibodies). Following staining, cells were incubated on ice for 30 min (4°C), protected from light. Cells were washed by adding 2 ml of DPBS per tube and centrifuging cells at 300 g for 5 min followed by supernatant aspiration. Cells were resuspended in 400 μl of PBS for flow cytometry analysis utilizing the BD LSR Fortessa Cell Analyzer (without UV).

FlowJo Version 10 (TreeStar, Ashland, OR) was used for data analysis. Relevant compensation matrices were constructed for each sample and applied to the data. Singlets were gated out initially by comparing SSC-W versus SSC-A for each sample. Next, multistain samples were compared to relevant FMO controls via quadrant gates to separate populations of cells and identify an upregulation of a specific marker as denoted in Table 2. Single-stained and unstained controls were used for further confirmation of gate borders. In this study, CD33+/CD80+ (pro-inflammatory) and CD33+/CD206+ (anti-inflammatory) populations were of principal interest for evaluating Mφ polarization.

TABLE 2.

Basal, pro-inflammatory, and anti-inflammatory Mφ markers for flow cytometry

	Marker
Cell phenotype	CD11b	CD33	CD80	CD206
Basal Mφ	+	+	−	−
Pro-inflammatory Mφ	+	+	+	−
Anti-inflammatory Mφ	+	+	−	+

Abbreviation: Mφ, macrophages.

2.12 |. Statistical analysis

A three-way ANOVA (with type III sums of squares) was used to compare the effect of three between-group factors, namely of hydrogel presence, culture type (VFF presence), and time (3, 24, 72 hr) upon cytokine production (TNF-α, IL-10) and Mφ viability respectively.

Separately, a further three-way ANOVA (with type III sums of squares) was used to compare the effect of three between-group factors, namely of cross-linker concentration (.005%, .01%, .02%), culture type (VFF presence), and time (3, 24 , 72 hr) upon cytokine production (TNF-α, IL-10) and Mφ viability respectively.

The normality of the distribution of cytokine production and viability were tested and a Box Cox transformation was applied on IL-10, TNF-α and viability variables in order to increase their normality for statistical testing.⁶² The residuals of the statistical models have been assessed ensuring that they were normally distributed and presented a homogenous variance. Post hoc comparisons using the Tukey HSD test were performed for further evaluations of significant differences. The criteria for significance were *p < .05, **p < .01, ***p < .001, and ****p < .0001. Analyses were carried out in R using the package “car” (Version 3.0–10).⁶³

3 |. RESULTS

3.1 |. Mechanical characterization

Twenty indentation tests were performed on each of the three-hydrogel samples. Data points were averaged to produce the mean Young’s modulus and SD of each hydrogel (Table 3). Results demonstrated that Young’s modulus increased with glyoxal cross-linker concentration. As the intended clinical use for this hydrogel would be in the vocal folds, for context, the mean Young’s modulus of in vivo human vocal folds at rest has previously been reported as approximately 12.66 kPa (2.45–29.4 kPa range).⁴⁵

TABLE 3.

The Young’s modulus of three glycol-chitosan hydrogels measured using atomic force microscopy indicated as the mean modulus across 20 measurements ± SD

Glyoxal concentration (%)	Young’s modulus (kPa)
0.005	1.11 ± 0.344
0.01	3.35 ± 0.790
0.02	7.86 ± 0.999

Shear plate rheometry was used to determine the storage modulus (G’) and loss modulus (G”) of the materials. All glycol-chitosan hydrogel samples tested demonstrated G’ values that remained larger than G” through the testing period (Figure 2). Results also indicated that the hydrogels tested had the capacity to elastically store energy and return, to an extent, to their initial configuration prior to mechanical force application. The stiffest hydrogel (0.02% glyoxal cross-linker) possessed the greatest elastic capabilities due to an increased level of cross-linking.

FIGURE 2.

The (a) storage modulus and (b) loss modulus of the three glyoxal cross-linked glycol-chitosan hydrogels assessed using shear plate rheometry. Glyoxal cross-linker concentrations used were 0.005, 0.01, and 0.02%

3.2 |. TNF-α cytokine secretion

3.2.1 |. All cultures: ANOVA for time by hydrogel presence by culture type

From the TNF-α ELISA results, a significant interaction effect was found (Table 4) between time and culture type (F [2, 58] = 4.66, p < .05) (Figure 3a). Further analyses indicated significant main effects: (1) time (F [2, 58] = 6.94, p < .01) and (2) hydrogel presence (F [1, 58] = 40.70, p < .0001) (Figure 3b). All other fixed effects and interactions were nonsignificant.

TABLE 4.

Results of the three-way ANOVA assessing the effect of time, hydrogel presence, and culture type on TNF-α detected. Bold indicates significant results

Source	Sum of squares	Df	Mean squares	F value	p
Time (T)	48.9	2	24.5	6.94	.002
Hydrogel presence (HP)	143.3	1	143.3	40.70	.000
Culture type (CT)	11.8	1	11.8	3.34	.073
T × HP	18.2	2	9.1	2.58	.084
T × CT	32.8	2	16.4	4.66	.013
HP × CT	7.8	1	7.8	2.21	.142
T × HP × CT	3.3	2	1.7	0.46	.631
Residuals	204.3	58	3.5

FIGURE 3.

(a) Interaction between culture type and Time on TNF-α detected (co-culture: N = 36; monoculture N = 36); (b) Main effect of hydrogel presence on TNF-α concentrations (hydrogel absent control group: N = 18; hydrogel present groups: N = 54). Data are graphed as concentration normalized to total protein levels (pg/g protein). The bars and error bars represent the mean and standard error of the data respectively. *p < .05, ***p < .001, ****p < .0001

Results from Tukey’s HDS indicated that overall TNF-α levels were higher for hydrogel cultures than for nonhydrogel cultures (p < .0001). While there was no significant difference between different time points for co-culture, TNF-α levels at 72 hr were significantly different from those at 3 hr (p < .001) and 24 hr (p < .05) for monoculture.

3.2.2 |. Hydrogel cultures: ANOVA for time by culture type by cross-linker concentration

Given the significant effect of hydrogel presence, a further ANOVA was performed for hydrogel cultures with cross-linker concentration included as a variable. Significant interaction effects (Table 5) were found for time by culture types by cross-linker concentration (F [4, 35] = 4.98, p < .01) (Figure 4), (2) time by culture type (F [2, 35] = 5.68, p < .01), (3) time by cross-linker concentration interaction (F [4, 35] = 16.55, p < .0001). Further analyses indicated significant main effects of: (1) time (F [2, 35] = 11.83, p < .0001), (2) cross-linker concentration (F [2, 35] = 14.88, p < .0001). All other fixed effects and interactions were nonsignificant. The simple effects from Tukey’s HDS tests indicated that overall TNF-α levels increased with time (3 ≃ 24 < 72 hr, p < .001) and cross-linker concentration (0.005 < 0.01 < 0.02%, p < .05).

TABLE 5.

Results of the three-way ANOVA assessing the effect of time, culture type, and cross-linker concentration on TNF-α detected. Bold indicates significant results

Source	Sum of squares	Df	Mean squares	F value	p
Time (T)	30.3	2	15.2	11.83	.000
Culture type (CT)	0.2	1	0.2	0.12	.727
Cross-linker concentration (CC)	38.1	2	19.05	14.88	.000
T × CT	14.5	2	7.25	5.68	.007
T × CC	84.7	4	21.18	16.55	.000
CT × CC	0.9	2	0.45	0.36	.070
T × CT × CC	25.5	4	6.4	4.98	.003
Residuals	44.8	35	1.3

FIGURE 4.

Three-way interaction between cross-linker concentration, Time and culture type on TNF-α detected in hydrogel cultures. Data is graphed as concentration normalized to total protein levels (pg/g protein). The bars and error bars represent the mean and standard error of the data respectively (N = 9 per co/monoculture group). *p < .05, **p < .01, ***p < .001, ****p < .0001

Further Tukey’s HDS tests indicated that the cross-linker concentration effect depended on both time and culture type (Table 6).

TABLE 6.

Post hoc comparisons summary for secreted TNF-α in hydrogel cultures. Bold indicates significant results

		Cross-linker concentration
Time	Culture type	0.005–0.01%	0.01–0.02%	0.005–0.02%
3 hr	(Mφ + VFF + hydrogel)	1.00 (n.s.)	−4.30 (p < .001)	−3.30 (p < .01)
3 hr	(Mφ + hydrogel)	1.57 (n.s.)	−3.40 (p < .01)	−1.83 (n.s.)
24 hr	(Mφ + VFF + hydrogel)	−2.61 (p < .05)	4.69 (p < .0001)	2.08 (n.s.)
24 hr	(Mφ + hydrogel)	−3.25 (p < .01)	1.24 (n.s.)	2.01 (n.s.)
72 hr	(Mφ + VFF + hydrogel)	−0.31 (n.s.)	−4.37 (p < .001)	−4.68 (p < .0001)
72 hr	(Mφ + hydrogel)	−2.16 (n.s.)	−0.67 (n.s.)	−2.83 (p < .05)

Abbreviations: Mφ, macrophages; VFFs, vocal fold fibroblasts.

At 3 hr, the general trend was similar for monoculture and co-culture: no significant difference between 0.005% and 0.01% concentrations and a significant difference between 0.01 and 0.02% concentrations for monoculture (p < .01) and co-culture (p < .001). TNF-α levels for 0.02% concentration were significantly higher than 0.005% concentration only for co-culture (p < .01).

At 24 hr, TNF-α levels were significantly higher for 0.01% than for 0.005% concentrations both for monoculture (p < .01) and co-culture (p < .05). However, while TNF-α levels were significantly lower for 0.02% concentrations than at 0.01% concentration for co-culture (p < .0001), there was no significant difference for monoculture between 0.01 and 0.02% concentrations.

At 72 hr, no significant difference was found between 0.005 and 0.01% concentration levels for either culture type. However, for 0.02% concentration, TNF-α levels in co-culture were significantly higher than for 0.01% concentration (p < .001). Levels for 0.02% were higher than0.005% both for co-culture (p < .0001) and monoculture (p < .05).

3.3 |. Il-10

3.3.1 |. All cultures: ANOVA for time by hydrogel presence by culture type

From the IL-10 ELISA results, significant interaction effects (Table 7) were found for: (1) time by hydrogel presence interaction (F [2, 58] = 11.45, p < .0001) (Figure 5(a)), and (2) time by culture type interaction (F [2, 58] = 3.30, p < .05) (Figure 5(b)). All main effects were found to be significant, namely: (1) time (F [2, 58] = 14.85, p < .0001), (2) hydrogel presence (F [1, 58] = 15.27, p < .0001), and (3) culture type (F [1, 58] = 7.16, p < .01). All other fixed effects and interactions were non-significant.

TABLE 7.

Results of the three-way ANOVA assessing the effect of time, hydrogel presence, and culture type on IL-10 detected. Bold indicates significant results

Source	Sum of squares	Df	Mean squares	F value	p
Time (T)	3,024.7	2	1,512.35	14.85	.000
Hydrogel presence (HP)	1,555.3	1	1,555.3	15.27	.000
Culture type (CT)	729.3	1	729.3	7.16	.009
T × HP	2,333.0	2	1,166.5	11.45	.000
T × CT	673.2	2	336.6	3.30	.044
HP × CT	299.6	1	299.6	2.94	.092
T × HP × CT	633.5	2	316.75	3.11	.052
Residuals	5,907.8	58	101.86

FIGURE 5.

(a) Interaction between hydrogel presence and time on IL-10 detected (Absent: N = 18; Present: N = 54); (b) Interaction between culture type and time on IL-10 detected (co-culture: N = 36; monoculture = 36). Data are graphed as concentration normalized to total protein levels (pg/g protein). The bars and error bars represent the mean and standard error of the data respectively. ***p < .001, ****p < .0001

Results from Tukey’s HDS indicated that overall IL-10 levels were not significantly different across time for nonhydrogel cultures. For hydrogel culture, IL-10 levels at 72 hr were significantly different than levels observed at 3 hr (p < .0001) and 24 hr (p < .0001). While there was no significant difference between different time points for monoculture, IL-10 levels at 72 hr were significantly different from those at 3 hr (p < .0001) and 24 hr (p < .001) for co-culture.

3.3.2 |. Hydrogel cultures: ANOVA for time by culture type by cross-linker concentration

Given the significant effect of hydrogel presence, a further ANOVA was performed for hydrogel cultures with cross-linker concentration included as a variable. Significant interaction effects (Table 8) were found: (1) time by culture type by cross-linker concentration (F [4, 35] = 3.60, p < .05) (Figure 6), (2) time by culture type (F [2, 35] = 21.99, p < .0001), (3) time by cross-linker concentration (F [4, 35] = 5.72, p < .01). Further analyses revealed main effects: (1) time (F [2, 35] = 80.65, p < .0001), (2) culture type (F [1, 35] = 33.45, p < .0001), (3) cross-linker concentration (F [2, 35] = 6.68, p < .01). All other fixed effects and interactions were nonsignificant.

TABLE 8.

Results of the three-way ANOVA assessing the effect of time, culture type, and cross-linker concentration on IL-10 detected. Bold indicates significant results

Source	Sum of squares	Df	Mean squares	F value	p
Time (T)	10,314.0	2	5,157.0	80.65	.000
Culture type (CT)	2,138.0	1	2,139.0	33.45	.000
Cross-linker concentration (CC)	855.0	2	427.0	6.68	.003
T × CT	2,813.0	2	1,406.0	21.99	.000
T × CC	1,464.0	4	366.0	5.72	.001
CT × CC	180.0	1	180.0	1.40	.259
T × CT × CC	920.0	4	230.0	3.60	.015
Residuals	2,238.0	35	64.0

FIGURE 6.

Three-way interaction between cross-linker concentration, time and culture type on IL-10 detected in hydrogel cultures. Data is graphed as concentration normalized to total protein levels (pg/g protein). The bars and error bars represent the mean and standard error of the data respectively (N = 9 per co/monoculture group). **p < .01, ***p < .001, ****p < .0001

Simple effects from Tukey’s HDS tests indicated that IL-10 levels increased with time (3 ≃ 24 hr and 24 < 72 hr, p < .001), cross-linker concentration (0.005% < 0.02%, p < .01), and were overall higher for co-culture ([Mφ + VFF + hydrogel] > [Mφ + hydrogel], p < .01). These simple effects were qualified by the interaction between culture type, time, and cross-linker concentration.

Finally, Tukey’s HDS tests revealed no significant difference between different levels of cross-linker concentration for both culture types at 3 hr and at 24 hr, nor at 72 hr for monocultures (Table 9). A significant difference between 0.005% and 0.01% cross-linker concentration (p < .01) and between 0.01% and 0.02% cross-linker concentration (p < .01) was found only at 72 hr for co-culture.

TABLE 9.

Post hoc comparisons summary for secreted IL-10 in hydrogel cultures. Bold indicates significant results

		Cross-linker concentration
Time	Culture type	0.005–0.01%	0.01–0.02%	0.005–0.02%
3 hr	(Mφ + VFF + hydrogel)	0.00 (n.s.)	0.00 (n.s.)	0.00 (n.s.)
3 hr	(Mφ + hydrogel)	0.71 (n.s.)	−5.36 (n.s.)	−4.65 (n.s.)
24 hr	(Mφ + VFF + hydrogel)	0.05 (n.s.)	3.52 (n.s.)	3.57 (n.s.)
24 hr	(Mφ + hydrogel)	0.65 (n.s.)	−2.20 (n.s.)	−1.54 (n.s.)
72 hr	(Mφ + VFF + hydrogel)	−21.21 (p < .01)	−25.79 (p < .001)	−46.97 (p < .0001)
72 hr	(Mφ + hydrogel)	−11.37 (n.s.)	1.29 (n.s.)	−10.09 (n.s.)

Abbreviations: Mφ, macrophages; VFFs, vocal fold fibroblasts.

3.4 |. Macrophage viability

Confocal images of Mφ stained with the viability assay are shown in Figure 7. Live/dead cell counts were statistically analyzed to determine differences across all cultures. As one data point ([Mφ + VFF + hydrogel], 0.005% cross-linker, 3 hr) was more than three standard deviations away from the mean it was considered an outlier and removed from the group analysis.

FIGURE 7.

Live/dead confocal images of Mφ on the underside of the basal membrane following culture with glyoxal cross-linked glycol-chitosan hydrogels. Panel A displays Mφ from (Mφ + VFF + hydrogel). Panel B displays Mφ from (Mφ + hydrogel). Live cells fluoresce green, dead cells fluoresce red. Cell density and distribution varied to an extent across the membrane surface and between samples. All images were taken at ×100 original magnification (scale bar = 100 μm). Mφ, macrophages; VFFs, vocal fold fibroblasts

3.4.1 |. All cultures: ANOVA for time by hydrogel presence by culture type

A significant interaction effect (Table 10) was found between hydrogel presence and culture type (F [1, 60] = 5.06, p < .05) (Figure 8). A main effect of hydrogel presence (F [1, 60] = 13.20, p < .0001) was indicated. All other fixed effects and interactions were nonsignificant. Results from Tukey’s HDS indicated that non-hydrogel cultures had higher viability than hydrogel cultures (p < .001). While there was no significant effect of culture type for nonhydrogel cultures, viability of co-culture was significantly higher than for monoculture for hydrogel cultures (p < .01).

TABLE 10.

Results of the three-way ANOVA assessing the effect of time, hydrogel presence, and culture type on Mφ viability. Bold indicates significant results

Source	Sum of squares	Df	Mean squares	F value	p
Time (T)	84.0	2	42.0	0.53	.590
Hydrogel presence (HP)	1,046.0	1	1,046.0	13.20	.000
Culture type (CT)	9.0	1	9.0	0.11	.741
T × HP	284.0	2	142.0	1.80	.175
T × CT	400.0	2	200.0	2.52	.089
HP × CT	401.0	1	401.0	5.06	.020
T × HP × CT	9.0	2	4.5	0.06	.944
Residuals	4,752.0	60	79.2

FIGURE 8.

Interaction between culture type and hydrogel presence for Mφ viability (%). The bars and error bars represent the mean Mφ viability and standard error of the data respectively (Absent: N = 18; Present: N = 54). **p < .01. Mφ, macrophages

A subsequent ANOVA for hydrogel cultures revealed no significant main effect or interaction (Supplementary Table 3). This result indicated that cross-linker concentration did not significantly influence viability.

3.5 |. Macrophage phenotyping

Flow cytometry was used to identify Mφ subpopulations in this study. The total percentage (%) of cells that displayed CD33+/CD80+ (pro-inflammatory) and CD33+/CD206+ (anti-inflammatory) expression are indicated in Table 11. The quadrant method applied to obtain populations is described in Figure 9. Overall, the population of CD33 +/206+ cells appeared more dominant than CD33+/80+ across all samples. The mean CD33+/206+ population for (Mφ + VFF + hydrogel) was consistently higher than (Mφ + hydrogel) across all hydrogel stiffness variations tested. The CD33+/80+ population remained relatively consistent across all hydrogel cultures regardless of hydrogel stiffness or VFF presence.

TABLE 11.

Mean % ± SD of CD33 +/CD80+ and CD33+/CD206+ cells in each culture group. Minimum number of events recorded per sample was 10,000

Culture	Cross-linker concentration	CD33+/CD80+ population (%)	CD33+/CD206+ population (%)
(Mφ + VFF + hydrogel)	0.005%	6.2 ± 1.1	90.2 ± 0.1
	0.01%	8.2 ± 6.1	87.0 ± 4.2
	0.02%	6.8 ± 1.5	81.7 ± 4.4
(Mφ + hydrogel)	0.005%	7.0 ± 0.7	68.0 ± 1.6
	0.01%	8.9 ± 2.9	61.0 ± 1.6
	0.02%	10.1 ± 6.7	77.9 ± 5.7
(Mφ)	N/A	4.7 ± 1.3	77.7 ± 13.8

Abbreviations: Mφ, macrophages; VFFs, vocal fold fibroblasts.

FIGURE 9.

Representative examples of the quadrant gates drawn to determine upregulations of (a) CD80 or (b) CD206 in comparison to relevant FMO controls. The example shown is for a hydrogel co-culture (0.005% glyoxal) sample. Quadrant names refer to whether that given population expressed none, one, or both markers. Numbers refer to the percentage of cells from the multistained samples only that were found in a given quadrant population. Contour plots (5% levels) in FlowJo were examined prior to gate drawing to assist with defining upregulated populations. Color key: Red = Multistain (A, B); Blue = FMO (80) (A) FMO (206) (B); Orange = FMO (33) (A, B). FMO, fluorescence minus one

4 |. DISCUSSION

The use of hydrogels to treat vocal fold scarring has the potential to improve patient therapeutic outcomes compared to current clinical practices.^1,64 When developing vocal fold hydrogels, it is necessary to establish the material composition that minimizes undesirable inflammation, maintains optimum mechanical integrity, and promotes beneficial immunomodulatory effects.^48,55 Hydrogel stiffness is a tunable material property that can be harnessed to modulate the immune response.^4,46,65,66 Hydrogel stiffness can be modified via the concentration of cross-linking agents applied.^14,53 However, chemical cross-linkers, such as glyoxal, are cytotoxic and thus defining an optimum concentration that provides mutual mechanical and immunomodulatory benefits is critical for regenerative biomaterial design.⁶⁷

In the current study, the concentration of glyoxal cross-linker applied to a glycol-chitosan hydrogel was investigated for its effect upon the inflammatory activity of Mφ. As increased hydrogel stiffness was acquired via increased glyoxal concentration, it was hypothesized that stiffer hydrogels would induce increased pro-inflammatory Mφ activity.¹⁴ Further, the presence of VFFs was hypothesized to promote anti-inflammatory Mφ activity via crosstalk between the cell types.³⁸ Results in this study indicated that hydrogel presence, hydrogel stiffness, and VFF presence all induced significant changes in Mφ inflammatory activity.

First, the ELISA results showed that hydrogel presence in cell cultures produced a significantly increased cytokine concentration of pro-inflammatory TNF-α (Figure 3) and anti-inflammatory IL-10 (Figure 4). The effect of hydrogel presence was anticipated because Mφ, as early responding immune cells, respond to foreign materials as part of the nonspecific inflammatory response that represents a major component of the innate immune response.⁶⁸ The chemical and physical properties of biomaterials strongly influence immune cell recruitment.^69–71 In the present study, the hydrogel components likely had a direct impact on the inflammatory activity that occurred. For example, the glyoxal cross-linker may have induced elevated pro-inflammatory activity given that it is a known cytotoxic material and has previously been associated with decreased VFF viability when used in high concentrations.^14,67 Cell death triggers an inflammatory response in the local environment that may have caused increased Mφ pro-inflammatory activity via the secretion of danger signals from dead cells that can stimulate inflammation.⁷² In fact, Mφ and VFFs have been identified as a major cell source of high-mobility group protein 1 (HMGB1), a typical danger signal cytokine, within the vocal fold mucosa.^28,36 An alternative source of increased pro-inflammatory activity could be the hydrogel’s chitosan polymer backbone. As a natural polymer, chitosan can include impurities within its polymer network.⁴⁸ Depending upon the nature of these impurities, they could significantly increase the immunogenicity of the hydrogel, thus increasing associated inflammatory activity.⁷³

Mφ viability was found to decrease with hydrogel presence (Figure 8). A possible explanation is that apoptosis typically occurs following inflammation resolution.⁷⁴ Hydrogel presence was responsible for inducing Mφ activation and inflammation. However, ultimately inflammation was resolved as suggested by the increase in IL-10 levels and large anti-inflammatory Mφ population detected via flow cytometry. Hence, greater Mφ apoptosis likely occurred following this resolution, leading to the significant difference found for Mφ viability between hydrogel presence and absence.

Second, hydrogel stiffness was associated with alterations in inflammatory activity in this study. Hydrogel stiffness has been demonstrated to influence cellular activity and can potentially be harnessed to direct the inflammatory response toward resolution.^6,50 The stiffness of the material cells that contact is important in facilitating their adhesion, contractility, and migration.⁷⁵ For the glycol-chitosan hydrogel, the stiffness ranged from 1.11 to 7.86 kPa (Table 3). Increased stiffness may lead to significantly increased levels of the pro-inflammatory cytokine, TNF-α (Figure 4). Interestingly, increased stiffness was also associated with significantly higher concentrations of the anti-inflammatory cytokine IL-10. Elevated IL-10 levels can promote healthy, nonfibrotic VFFs and thus modulate Mφ toward inflammatory resolution.^19,38 The inherent plasticity of Mφ enables them to switch between pro- and anti-inflammatory phenotypes when polarized by external stimuli.⁷⁶ Hence local Mφ populations can include coexistent pro- or anti-inflammatory Mφ subpopulations. Flow cytometry results, collected at the 72 hr time-point, indicated the presence of such coexistent Mφ subpopulations. Notably, the presence of both pro- and anti-inflammatory Mφ for the stiffest hydrogel could explain its association with peak levels for both the pro- and anti-inflammatory cytokines assessed.

Stiffer hydrogels have also been associated with modulating fibroblast activities including extended focal adhesions, spreading, and a well-defined cytoskeleton.^8,9,52 In contrast, soft matrices were shown to induce cells to display increased TNF receptor clustering associated with NF-κB pathway activation.⁷⁷ The NF-κB pathway is related to the production of excessive pro-inflammatory cytokines, such as TNF-α that induces fibroblasts to overproduce ECM proteins associated with fibrosis.^78,79 Fibroblasts may have the capacity to tune their internal stiffness to that of the surrounding substrate as a potential mechanism for directing localized inflammatory activities.⁸⁰ Such fibroblast activity may be another possible cause of the increased cytokine concentrations detected for the stiffest hydrogels in this study.

Third, the presence of VFFs in the model was associated with increased anti-inflammatory activity compared to controls. Throughout the inflammatory response to a foreign material, such as the hydrogel presented, Mφ and fibroblasts communicate constantly via autocrine and paracrine signaling.⁶⁰ The release of pro-inflammatory factors from Mφ, including TNF-α, influences the growth, migration, and activity of fibroblasts and ultimately reinforces the local inflammatory response.³⁵ TNF-α also has involvement in the stimulation of MMP secretion from Mφ and VFFs in vocal fold tissue.^64,81,82 Results from the ELISA assessment of TNF-α indicated that the presence of VFFs was associated with lower TNF-α levels compared to monoculture controls as time increased (Figure 4). This finding agreed with literature elsewhere that VFF may suppress Mφ pro-inflammatory cytokine production and regulate inflammatory activity via toll-like receptor 4 (TLR4).^38,83

IL-10 typically peaks after the initial pro-inflammatory response and is associated with inflammatory resolution.⁵ In this study, IL-10 production was delayed compared to pro-inflammatory TNF-α for all cultures assessed, with minimal IL-10 detected prior 72 hr. However, at 72 hr a significant increase in IL-10 was detected for (Mφ + VFF + hydrogel) compared to (Mφ + hydrogel) (Figure 6), potentially driven by the VFFs modulating Mφ inflammatory activity.³⁸ Flow cytometry findings also suggested that VFF presence produced a greater proportion of anti-inflammatory (CD33+/CD206+) Mφ compared to respective monoculture controls.

4.1 |. Limitations and future directions

Mφ and fibroblasts, as two principal effector cell types of the initial inflammatory response to biomaterials, were included in this study. The incorporation of other immune cells, such as neutrophils or T-cells, should be considered for further hydrogel evaluation. As an alternative to primary cells, Mφ and VFF cell lines were used in this study to ensure a sufficient supply of cells for the multiple time-points and measurements included in the evaluation. However, despite the advantages of immortalized cell lines, such as ease of manipulation and expansion, significant drawbacks are attached to their use. For example, THP-1 Mφ have an inability to mimic the full spectrum of human Mφ phenotypes and are restricted to simplified models investigating polarization and functional implications.^61,84 Future work with the glycol-chitosan hydrogel should attempt to replicate inflammatory trends observed in this study while using primary cells or small-animal in vivo models.

In this study, a transwell system was applied to study the paracrine signaling in co-cultures of Mφ and VFFs in response to the glycol-chitosan hydrogel. Mφ were seeded on the basolateral transwell membrane while VFFs were cultured within the hydrogel of the upper well. However, the membrane of the transwell insert presented notable technical challenges in studying the interaction of Mφ with the hydrogel surface and characterizing VFFs with confocal microscopy. Future in vitro investigations with more advanced 3D culture models, such as organoids, may help circumvent aforesaid issues related to the transwell co-culture system.

Last, cross-linker concentration not only varies hydrogel stiffness but also other material properties such as pore size and degradation profile. As these materials properties are often dependent, decoupling them is experimentally challenging. Work is ongoing to develop a computational model of biomaterials to decipher how individual and combined material properties influence cell activities and long-term tissue growth.

5 |. CONCLUSION

Increased hydrogel stiffness modulated Mφ behavior toward inflammatory resolution as shown by the heightened levels of IL-10. The presence of VFFs within hydrogel cultures further suppressed pro-inflammatory activity as shown by significant changes in cytokine concentrations and the dominant anti-inflammatory macrophage population detected. An in vitro evaluation of material-related inflammation that considers hydrogel physical properties and crosstalk between Mφ and VFFs is recommended for use during the determination of optimum vocal fold hydrogel characteristics from the immunological perspective.